1. Field of the Invention
The present invention relates to methods for handling messages after reset and reestablishment in wireless systems, and more particularly, to a method for handling radio resource control (RRC) radio bearer 3 (RB3) messages after re-establishment due to SRNS relocation in an acknowledged mode (AM) radio link control (RLC) entity that belongs to RB3.
2. Description of the Prior Art
In a universal mobile telecommunications system (UMTS), a universal terrestrial radio access network (UTRAN) communicates with a plurality of mobile stations, also referred to as user equipment (UE). FIG. 1 is a diagram illustrating network layers of the UMTS protocol. Layers of a UE 10 and a UTRAN 20 are shown. Both the UE 10 and the UTRAN 20 are illustrated having the same network layers. Level 3 layers contain both a packet data convergence protocol (PDCP) layer and a radio resource control (RRC) layer. Level 2 layers contain both a radio link control (RLC) layer and a media access control (MAC) layer. Level 1 contains a physical layer. These layers are all well known to those skilled in the art, and will only be described as they relate to the present invention.
The present invention primarily concerns interaction between the PDCP layer and the RLC layer, as well as interaction between the RRC layer and the RLC layer. The terms service data unit (SDU) and protocol data unit (PDU) are well known in the art. The terms PDU and SDU are relative with respect to the current layer being described. An SDU is a piece of information received from a layer above the current layer for transmission using the service of the current layer. A PDU is a piece of information processed by the current layer. Please refer to FIG. 2. FIG. 2 is a diagram illustrating PDCP PDUs 30, 32 being segmented into RLC PDUs 34, 36, 38 according to the prior art. The RLC receives PDCP PDUs as RLC SDUs and segments the RLC SDUs into RLC PDUs. For simplicity, packet headers are ignored in this diagram. The size of each RLC PDU 34, 36, 38 is configured by the UTRAN 20, whereas the size of each PDCP PDU 30, 32 is not set by the UTRAN 20. As shown in FIG. 2, a first PDCP PDU 30 is segmented to produce RLC PDUs 34 and 36 and part of RLC PDU 38. Since there is still room in the RLC PDU 38 for more data, part of the PDCP PDU 32 data is also added to the RLC PDU 38. This segmentation process continues for dividing all PDCP PDUs into RLC PDUs.
In acknowledged mode (AM) transmission and reception, each time an AM RLC entity sends a PDU, a corresponding acknowledgement should be received from the AM RLC entity receiving the PDU. For radio bearers that are configured to support lossless SRNS relocation, when the SRNS relocation is performed, the transmitting side of the AM RLC entity is re-established. Further, for radio bearers configured to support lossless downlink RLC PDU size change, occasionally the UTRAN 20 will change the size of the downlink RLC PDU segment size. In this case, upper layers will also request the transmitting side of the UTRAN AM RLC entity to be re-established. Please refer to FIG. 3. FIG. 3 illustrates a situation in which the UTRAN AM RLC entity is re-established. The UTRAN PDCP entity submits PDCP data PDUs 40-43 with sequence numbers (SNs) equaling 77, 78, 79, and 80 to the UTRAN AM RLC entity. The UTRAN AM RLC entity then begins transmitting the PDCP data PDUs to the UE 10. When the SRNS relocation is performed, or when the downlink RLC PDU size is changed by the UTRAN 20, the transmitting side of the UTRAN AM RLC entity is re-established. At this time, the PDCP Data PDUs 40, 41 with SNs=77 and 78 have been transmitted by the transmitting side of UTRAN AM RLC entity but have not been acknowledged positively. According to the prior art, when the transmitting side of the UTRAN AM RLC entity is re-established, the AM RLC entity discards all RLC SDUs that have been transmitted completely in the transmitting side. This means that RLC SDUs containing data for the PDCP Data PDUs 40, 41 with SN=77 and 78 are discarded. At this point the PDCP Data PDUs 42, 43 with SNs=79 and 80 have not been transmitted by the transmitting side of UTRAN AM RLC entity. These PDCP Data PDUs 42, 43 are put in a buffer for later transmission.
Due to the re-establishment, the AM RLC entity re-segments the SDUs that were not discarded into AM data (AMD) PDUs with the configured RLC PDU size (which may be different from the size before the re-establishment) so an RLC SDU containing PDCP Data PDU 42 with SN=79 and an RLC PDU containing PDCP PDU 43 with SN=80 are re-segmented. If the PDCP entity in the UTRAN 20 has to trigger a PDCP SN synchronization procedure, it submits one PDCP SeqNum PDU 44 to lower layers. Because the smallest unacknowledged SN is 77, the PDCP entity submits a PDCP SeqNum PDU 44 with SN=77 containing the same data as the PDCP Data PDU 40 with SN=77. PDCP Data PDUs 45, 46, 47 with SN=78, 79, and 80 are also submitted to the AM RLC entity again.
Unfortunately, since the PDCP Data PDUs 42, 43 with SNs=79 and 80 were not discarded during re-establishment, they were put in a buffer for later transmission. That means when the UTRAN AM RLC entity later submits PDCP SeqNum PDU 44 with SN=77 and PDCP Data PDUs 45, 46, 47 with SN=78, 79, and 80, PDCP data PDUs with SN=79 and 80 are transmitted twice. Not only does this waste radio resources by transmitting PDUs twice, it also disrupts the delivery sequence of the PDCP data PDUs since the SNs=79 and 80 are sent from the buffer before the SNs=77-80 are sent after re-establishment.
Please continue to refer to FIG. 3. The same problem described above that applies to re-establishment also applies to RLC reset events. Consider again the situation for a radio bearer that is either configured to support the SRNS relocation or configured to support lossless downlink RLC PDU size change. The UTRAN PDCP entity submits PDCP data PDUs 40-43 with SNs equaling 77, 78, 79, and 80 to the UTRAN AM RLC entity. The UTRAN AM RLC entity then begins transmitting the PDCP data PDUs to the UE 10. If a condition of RLC reset is fulfilled, the RLC reset procedure is triggered. At this time, the PDCP Data PDUs 40, 41 with SNs=77 and 78 have been transmitted by the transmitting side of UTRAN AM RLC entity but have not been acknowledged positively. According to the prior art, when the UTRAN AM RLC entity is reset, the AM RLC entity discards all RLC SDUs that have been transmitted completely in the transmitting side. This means that RLC SDUs containing data for the PDCP Data PDUs 40, 41 with SN=77 and 78 are discarded. At this point the PDCP Data PDUs 42, 43 with SNs=79 and 80 have not been transmitted by the transmitting side of UTRAN AM RLC entity. These PDCP Data PDUs 42, 43 are put in a buffer for later transmission.
Next, the PDCP entity in the UTRAN 20 triggers a PDCP SN synchronization procedure by submitting one PDCP SeqNum PDU 44 to lower layers. Because the smallest unacknowledged SN is 77, the PDCP entity submits a PDCP SeqNum PDU 44 with SN=77 containing the same data as the PDCP Data PDU 40 with SN=77. PDCP Data PDUs 45, 46, 47 with SN=78, 79, and 80 are also submitted to the AM RLC entity again.
Unfortunately, no handling method is specified for the RLC SDUs that were not transmitted before the reset in the transmitting side of the UTRAN AM RLC entity. If the PDCP Data PDUs 42, 43 with SNs=79 and 80 were not discarded during the reset, they were put in a buffer for later transmission. That means when the UTRAN AM RLC entity later submits PDCP SeqNum PDU 44 with SN=77 and PDCP Data PDUs 45, 46, 47 with SN=78, 79, and 80, PDCP data PDUs with SN=79 and 80 are transmitted twice. Not only does this waste radio resources by transmitting PDUs twice, it also disrupts the delivery sequence of the PDCP data PDUs since the SNs=79 and 80 are sent from the buffer before the SNs=77-80 are sent after the reset.
Please refer to FIG. 4. FIG. 4 illustrates a situation in which the UTRAN AM RLC entity belonging to RB3 is re-established. Consider RRC signaling radio bearer RB3. The RRC submits RB3 messages 1 and 2 50, 51 to the RLC. When the RB3 RLC is re-established due to SRNS relocation, the RRC RB3 message 1 50 has been transmitted but is not positively acknowledged. RRC RB3 message 2 51 has not been transmitted. After RLC re-establishment, the RRC retransmits the RB3 messages 1 and 2 52, 53. However, the RB3 RLC also keeps the RRC RB3 message 2, and may keep the RRC RB3 message 1. Therefore, the RRC RB3 messages 1 and 2 are duplicated in the buffer, and will be transmitted twice. This causes unnecessary retransmissions, and wastes radio resources.
Likewise, FIG. 4 also illustrates a situation in which the UE AM RLC entity belonging to RB3 is re-established. Consider RRC signaling radio bearer RB3. The RRC submits RB3 messages 1 and 2 50, 51 to the RLC. When the RB3 RLC is re-established due to SRNS relocation, the RRC RB3 message 1 50 has been transmitted, but is not positively acknowledged. RRC RB3 message 2 51 has not been transmitted. After RLC re-establishment, the RRC retransmits the RB3 messages 1 and 2 52, 53. However, the RB3 RLC also keeps the RRC RB3 message 2, and may keep the RRC RB3 message 1. Therefore, the RRC RB3 messages 1 and 2 are duplicated in the buffer, and will be transmitted twice. This causes unnecessary retransmissions, and wastes radio resources.
The prior art also suffers from other problems because the proper handling for certain situations is not specified. For instance, the lossless function is not provided for uplink transmission in a UE PDCP entity for a radio bearer configured to support lossless DL RLC PDU size change. The handling for RLC SDUs that have not been transmitted completely before RLC re-establishment is not specified clearly in the specification. Therefore, a bad design, such as discarding RLC SDUs that have not been transmitted completely before the RLC re-establishment, will introduce transmission delays because the discarded RLC SDUs have to be recovered by upper layer retransmission.
Similarly, the handling for RLC SDUs that have not been transmitted completely before RLC reset is not specified clearly in the specification. Therefore, a bad design, such as discarding RLC SDUs that have not been transmitted completely before the RLC reset, will introduce transmission delays because the discarded RLC SDUs have to be recovered by upper layer retransmission.